Before the Joint Committee on Health Care
Financing
October 3, 2011
The Massachusetts Medical Society wishes to be recorded in
strong support for Senate Bill 494. Rising health care
costs and further attention to the quality of health care delivered
in the Commonwealth have driven payers, employers, and providers to
seek out new ways to measure costs and value with regard to the
provision of health care. Currently, there are a number of
initiatives underway at the state and federal level that seek to
achieve the above stated goals. To say that they are
experimental would be an understatement. To say their
ramifications are magnified by the current practice environment for
physicians in Massachusetts would be an even greater understatement
as the administrative burden is substantially increased as multiple
measurement tools are applied to the same group of
physicians.
To assist in the successful development and implementation of
performance measurement, reporting and rewards programs, The
Medical Society has done research with national experts to develop
recommendations and guidelines to help ensure that methodologies
used to develop these programs are carefully and thoughtfully
researched, developed and validated with the goal of protecting
patients from exposure to potential inaccuracies and unintended
consequences. We believe that particular attention must be
paid to transparency regarding all aspects of the methodologies and
criteria used to judge physicians. We equally believe that
physician involvement is needed to define both valid and meaningful
quality and cost measures as well as an appropriate methodology to
use quality and cost information. In addition, physician
involvement is necessary in determining a process to correct
inaccuracies in the data. To these ends we offer the
following comments:
Consumer directed programs that include "preferred networks"
should:
- Support the Patient/Provider Relationship:
Quality and cost efficiency measurement programs should be directed
at supporting and improving patient-physician relationships.
They should not create obstacles for providers treating patients
regardless of their health condition, ethnicity, economic
circumstance, demographics, or treatment compliance pattern.
- Support Sound Performance Measures: Quality,
efficiency and cost performance measures should be evidence-based,
valid, reliable, broadly accepted, and clinically meaningful.
Measures should be consistent with those collected by national or
regional organizations such as the AMA's Physician Performance
Consortium, Ambulatory Care Quality Alliance (Better Quality
Initiative), National Quality Foundation, JCAHO, Massachusetts
Health Quality Partners (MHQP), and the Centers for Medicare and
Medicaid Services thus facilitating an alignment of measurement
goals in the marketplace.
- Evidence-based quality and cost measures should be evaluated in
relation to each other. Measures should result in no
unintended harmful consequences.
- Support Methodology/Data Transparency:
Complete descriptions of all criteria, algorithms, methodologies
and data sources used in such programs should be made readily
available to plan members and participating physicians, as should
all of the underlying individual physician quality, cost,
efficiency and patient satisfaction data.
- There should be a statistically valid reason for judging any
data used and arbitrary cut-offs must be avoided; and physicians
whose practices are too new, too small to measure, or different
from their peers should be handled separately from
others.
- Support Data Accuracy: Measurements
should be accurate and timely. Physicians should be given
patient-level drilldowns for the efficiency measure, and patient
lists for the quality measures. There should be a formal
feedback and correction mechanism so that errors uncovered by
physicians, plans, and other analysts can contribute to improving
the evaluation system.
- Data should be adjusted for such items as sample size and
case-mix composition, outliers when appropriate, socio-economic
differences when possible, appropriate use of preventive care and
other under-utilized interventions; reasonable targets should be
set for each measure; and adjusted to account for variations in the
cost of delivering care which are outside the providers' control
(e.g., variations in payor mix, area wage index, and state mandated
requirements)
- Data should not be attributed to an individual physician
unless limited to the results of the diagnoses the individual
physician has made and the care he/she has provided. If the data
reflects the results of all of the care received by a cohort of
patients (rather than just the care provided/ordered by an
individual physician), the results should be attributed to a
physician practice or network only - with no individual physician
attribution.
- Measurements should be at a group level until data accuracy is
improved and the methodology is further validated.
- Support Data Sharing at Individual Level for Quality
Improvement: : At this time, because of limitations
to the current system and issues surrounding attribution, and
appropriate volume it is inappropriate to judge physician
performance at a level finer than the large group level. This
would include integrated health care systems and Independent
Physician Associations (IPAs). It is, however, appropriate
and desirable to provide data to medical groups and physicians that
can be drilled down to the individual physician and individual
patient for purposes of providing best care and improving the
process of measurement.
- Support Physician Involvement in the Process:
Practicing physicians, hospitals, and their professional
organizations should be involved in the design and ongoing
modification of programs such as these that judge physicians in
order to be fair to physicians and physician patient relationships.
Results must be shared with physicians well in advance of any final
judgments about their performance. Criteria used to judge
physician's performance should be circulated well in advance of any
final opinions about their performance. There should not be any
introduction of unnecessary administrative complications to
practices. Physicians should be provided with specific behaviors
(action items) by which they can improve their results. A
uniform approach to measurements should be adopted.
- Support a Uniform Format for Reporting: In
order for physicians and patients to better understand and make use
of the information available, information about physicians should
be provided in a common format.
The Group Insurance Commission has launched an initiative
referred to as the Clinical Performance Improvement (CPI)
Initiative. This initiative involves the tiering of
physicians which in the opinion of the Medical Society and others
must involve a very high level of rigor in the development and use
of the metrics upon which the tiering system is constructed.
Due to the broad based nature of the GIC's initiative and the
serious nature of the ramifications if not done correctly, the MMS
engaged four national consultants to review the CPI
initiatives. A copy of the Executive Summary of that
report is included. As was noted in the report issued by the
consultants, "physicians' quality and cost efficiency profiles are
developed from the health care claims databases of participating
health plans. Data accuracy is critical to performance
evaluations for all stakeholders, decreases the risk of unintended
consequences, and increases physician distrust of the
system." The legislation before you is an attempt
to codify the recommendations of the consultants in order to more
effectively promote quality improvement and appropriate cost
control while not adversely impacting the patient physician
relation or the physician practice environment in
Massachusetts.
Lastly, we believe that the Legislature recognized the
importance of thoughtful performance benchmarks in the area of cost
containment and quality healthcare in Chapter 58 of the Acts of
2006. In Section 3 of the law, the Legislature added a new
section 16L to Chapter 6A of the General Laws. Subsection (g)
states "that performance benchmarks shall be clinically important
and include both process and outcome data, shall be standardized,
timely, and allow and encourage physicians, hospitals and other
health care professionals to improve their quality of care.
Any data reported by the council should be accurate and
evidence-based and not imply distinctions where caparisons are not
statistically significant." S. 494 further refines this
goal.
For all of the reasons stated above, we urge your favorable
report on S. 494.
